Growing up in a digital environment: The role of digital media use in European children’s and adolescents’ health and health behaviours

Abstract

Background Children’s and adolescents’ health state suffers from the double burden of metabolic and mental health disorders, representing a critical public health matter. As never before, today’s children are growing up in a saturated digital media (DM) environment. Despite the immense opportunities for learning and self-development, little is known about the role of DM exposure on children’s health. Aim This doctoral dissertation aims to provide evidence on the potential association of DM exposure with health outcomes, including metabolic syndrome and cognitive functioning, as well as health behaviours, namely dietary intake, eating habits, and sensory taste preferences in children and adolescents. Methods The present cumulative thesis is constituted of four papers: one systematic literature review (SLR, paper 1) and three original investigations (papers 2, 3, and 4). In paper 1, a total of 35 studies conducted worldwide were reviewed, critically appraised, and synthesized. These studies examined the association of social media (SM) exposure with the dietary intake, breakfast skipping, and nutrition literacy of healthy children and adolescents. The SLR was based on the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) guidelines. The empirical research conducted in papers 2 to 4 used data provided by children aged 2-18 years of IDEFICS/I.Family cohort. The cohort was carried out in three waves across nine European countries. The first examination wave (i.e., baseline, W1) was conducted during 2007-2008, and 16,229 children participated. The second examination wave (i.e., first follow-up, W2) was conducted during 2009-2010 and included 13,596 children. The third examination wave (i.e., second follow-up, W3) was conducted during 2013-2014 and included 9,617 children and adolescents. The overarching aim of the cohort was to identify dietary- and lifestyle-induced health effects in children and adolescents, considering sensitive developmental periods, and to develop a community-based intervention on childhood obesity. Across paper 2 to paper 4, the analysis group varied from 3,261 to 10,359 participants after respective inclusion/exclusion criteria were met. DM exposure (hours/day) was self-reported, including: i) television viewing (TV), ii) computer/game console (PC), iii) smartphone, and iv) internet exposure. The related behaviour of media multitasking, defined as the simultaneous use of several media, was also reported. In paper 2, sensory taste preferences for sweet, fatty, salty, and bitter taste were evaluated via a Food and Beverage Preference Questionnaire. In paper 3, measures of cognitive functioning, namely cognitive inflexibility and decision-making ability were assessed via computerized tests, while emotion-driven impulsiveness was self-reported. In paper 4, metabolic syndrome (MetS) and its components: abdominal obesity (via waist circumference), blood pressure (systolic and diastolic), insulin resistance (homeostasis model assessment for insulin resistance (HOMA-IR) or fasting glucose), and dyslipidaemia (HDL-cholesterol and triglycerides), were objectively measured. Age and sex-specific z-scores and monitoring levels (≥90th percentile, as defined by Ahrens et al. 2014) were considered for each metabolic outcome. The statistical approach used to investigate the associations of interest varied depending on the research questions. Logistic regression models were used to examine associations between exposures and dichotomized outcomes (papers 2 and 3). Latent class analyses were performed to identify underlying patterns of DM exposure (paper 3), based on a combination of using the individual media (in categories). In paper 4, to examine the longitudinal association of DM exposure with MetS and its components, a two-step trajectory approach was used: first, the age-dependent trajectories of DM exposure were calculated using linear mixed models; second, to estimate the association between childhood DM trajectory and MetS at follow-up, generalized linear mixed models were used. Across papers, analyses were stratified by sex, age, country of residence, parental educational status, and family structure, to characterize children and adolescents that are most vulnerable to the potential negative impact of DM exposure. Results The SLR revealed a dose-dependent relationship between SM exposure and daily intake of sugar and caffeine and the consumption frequency of sugar-sweetened beverages in both children and adolescents. SM exposure was also associated with low frequency intake of fruits and vegetables and less frequent breakfast consumption. No association between SM exposure and nutrition literacy was observed. SM exposure, measured either as WhatsApp use, watching YouTube videos, or exposure to SM influencer’s advertising on Instagram, led to an increased intake of unhealthy food and beverages at ad-libitum and after two years. A neuro-physiological mechanism was identified: exposure to digital food images increased the neural activation of brain areas related to reward and attention. Peers’ presence on SM but not of SM influencers, showed a potential to improve adolescents’ vegetable intake. In IDEFICS/I.Family children exposure to DM increased over age, from 2.4 h/day at the age of two years to 5.5 h/day at the age of 16 years. This increase was steeper among boys compared to girls. Country differences were also observed, where Estonian, Cypriot, and Swedish children had the highest DM increase, while Spanish children showed the lowest DM increase. The observational research conducted in paper 2 showed that prolonged DM exposure (>2 h/day) was associated with a high preference score for sweet, fatty, and salty-tasting foods among adolescents, especially females. An inverse association between prolonged DM exposure and bitter taste preference was observed among males. In paper 3, it was observed that one additional hour of exposure to smartphones and the internet, and higher media multitasking was positively associated with children’s emotion-driven impulsiveness and cognitive inflexibility and negatively associated with decision-making ability. Compared to participants with low exposure to all media, participants with “high smartphone and internet, in combination with medium TV and low PC exposure”, showed higher scores for emotion-driven impulsiveness and cognitive inflexibility and a lower score for decision-making ability. In paper 4, it was found that increasing DM exposure during childhood was positively associated with the z-scores of MetS, waist-circumference, HOMA-IR, HDL-c-1, and triglycerides after two or six years. The stratified analyses revealed that associations were independent of moderate to vigorous physical activity. Children with an above-average DM increase over age (DM slope> 0 h/day/year) showed a 22% higher risk for later incident MetS. This risk was higher among boys compared to girls (41% and 10%, respectively). Conclusion The findings of this thesis suggest that DM exposure is associated with unfavorable dietary intake and poor eating behaviors. A neuro-physiological mechanism and a clear impact of peers and SM influencers on the SM environment explain these findings. The results also showed that DM exposure is positively associated with a preference for sweet, fatty, and salty-tasting foods and negatively associated with a preference for bitter-tasting foods. This suggests that DM exposure may lead to obesity by favoring the taste preference of unhealthy foods over healthy ones. Moreover, exposure to modern DM was positively associated with children’s emotion-driven impulsiveness and cognitive inflexibility and inversely associated with decision-making ability. This sheds light on a new potential mechanism by which DM exposure leads to poor mental health in children and adolescents. Finally, the findings support the hypothesis that increasing DM exposure during childhood may be an independent risk factor for metabolic syndrome later in life, with boys being at higher risk. These long-term associations need to be confirmed in other populations of children and adolescents, considering not only duration but also patterns of DM exposure, as well as children with an unfavorable background regarding socio-economic status, learning difficulties, or predisposing mental disorders. Further interdisciplinary, longitudinal studies may consider the interplay between health determinants in the physical and digital environment to identify potential intervening factors to promote children’s health in a hybrid world. Future health interventions may consider a precautionary approach and use the identified mechanisms to increase children’s and adolescents’ resilience against the potential adverse health effects of the digital environment at an early stage of their development.